DISCORDANT BIRTH-WEIGHT AND LATE FETAL DEATH IN LIKE-SEXED AND UNLIKE-SEXED TWIN PAIRS - A POPULATION-BASED STUDY

Objective To study the relation between pregnancy loss (one or two fet al deaths in a pair) in twin pregnancy and gender (like-sexed and unli ke-sexed pairs), discordant birthweight, and placentation. Design The type of placentation for gestations with pregnancy loss was studied, b ased on material retrieved from the original medical records, for a de fined region comprising 20.0 % of the total twin population. Subjects and setting Four thousand one hundred and ninety-one unlike-sexed and 10875 like-sexed twin pairs born in Sweden between 1973 and 1989, in m ost cases with a gestational duration 28 or more completed weeks. Resu lts The relative risk (RR) for pregnancy loss in like-sexed pairs, com pared with unlike-sexed ones, was 23 and the 95% confidence limits (CL ) 1.7-3.1. Not until discordance exceeded 999 g did the pregnancy loss rate for unlike-sexed pairs increase significantly, compared with the (first) stratum with discordance less than 250 g (RR = 6.3; CL 3.5-11 .3). For like-sexed twin pairs a higher pregnancy loss rate, compared with the first stratum, was seen already in the stratum with discordan ce 250-499 g (RR = 1.3; CL 1.0-1.8); a significant increase in pregnan cy loss rate was found in all the three strata 500-749 g (RR = 2.1; CL 1.5-3.0), 750-999 g (RR = 3.5; CL 2.3-3.0), and 1000 g or more (RR = 10.9; CL 8.4-14.2), respectively. When calculating the discordance as a percentage of the weight of the larger twin, unlike-sexed pairs expe rienced a significant increase in pregnancy loss when discordance exce eded 40 to 50 %. For like-sexed ones, the corresponding figures were 2 0 to 30 %. In the 47 unlike-sexed pregnancies complicated by pregnancy loss, both twins died in two pregnancies (43 %), whereas for like-sex ed pairs the corresponding figures were 65/279 (23.3 %). Of 47 gestati ons with pregnancy loss in a defined region, 32 were monochorionic (mo nozygous), nine were like-sexed dichorionic (monozygous or dizygous), and six were unlike-sexed (dizygous). Conclusions Pregnancy loss was t wice as high in like-sexed compared with unlike-sexed pairs, and only in like-sexed pairs was pregnancy loss strongly correlated to birthwei ght discordance. In twin pregnancies with one fetal death the risk for the surviving twin to succumb is five to six times higher in like-sex ed compared with unlike-sexed pairs and is most probably related to mo nochorionicity.